1. Field of the Invention
The present invention is directed to methods for upgrading bandwidth in a fiber optic system. More specifically, the present invention is directed to methods for upgrading bandwidth in a fiber optic system that utilizes Raman amplification.
2. Description of the Related Art
To increase the bandwidth of optical fibers, two key amplification technologies have been developed and used by the telecommunication industry: erbium-doping and stimulated Raman scattering. While both of these technologies have proven useful in increasing band width, there are shortcomings associated with each.
Silica-based optical fiber has a lowest loss window around 1550 nm with approximately 25 THz of bandwidth between 1430 and 1620 nm. In this wavelength region, erbium-doped fiber amplifiers (EDFAs) are widely used. However, the absorption band of an EDFA nearly overlaps its emission band. For wavelengths shorter than about 1525 nm, erbium-atoms in typical glasses will absorb more than amplify. Consequently, the full bandwidth potential for silica-based optical fibers have not been attained. To broaden the gain spectra of EDFAs, the silica core of the optical fibers have also been doped with aluminum or phosphorus to broaden the emission spectrum. Nevertheless, the absorption peak for the various glasses is still approximately 1530 nm, which narrows the usable bandwidth. As discussed in the prior art section of U.S. Pat. No. 6,101,024 to Islam et al., two-band architecture for an ultra-wideband EDFA has been proposed with an optical bandwidth of 80 nm. To obtain a low noise figure and high output power, the two bands share a common first gain section and have distinct second gain sections. In particular, the 80 nm wavelength bandwidth comes from one amplifier (for conventional band or xe2x80x9cC-bandxe2x80x9d) from 1525.6 to 1562.5 nm, and another amplifier (for long band or xe2x80x9cL-bandxe2x80x9d) from 1569.4 to 1612.8 nm. In another proposed system, a 54 nm gain bandwidth is achieved with two EDFAs in a parallel configuration, i.e., one optimized for 1530-1560 nm and the other optimized for 1576-1600 nm. However, such systems require the use of additional optical repeaters to achieve the multiple wavelength transmissions. Thus, as can be readily appreciated by one of ordinary skilled in the art, such an upgrade method would require a significant capital investment and would also render different nodes of the network inoperable at different times thereby potentially disrupting service to many subscribers.
Raman amplification is disclosed in U.S. Pat. No. 5,959,750 to Eskildsen et al. for increasing transmission capacity. Here, amplification occurs directly in the transmission fibers of the fiber optic system itself For a system that relies entirely on Raman amplification, the need for specially doped fibers is obviated. Thus, the use of Raman gain as taught by Eskildsen et al. facilitates the upgrade of a transmission system without significant cost. This use of Raman amplification is referred to as distributed Raman-assisted transmission (D-RAT), and is implemented via counter-propagation (with respect to the signals) of pump light which is downshifted by approximately 100 nm from the signal wavelength band. In addition to lower cost, D-RAT offers the benefit of lower accumulated amplified spontaneous emission (ASE) noise build-up and lower launch power for decreased fiber nonlinearities. So D-RAT acts as a low noise, pre-amplifier before each in-line EDFA in a long-haul transmission link to improve system performance (as measured by Q) by several decibels (dB). The Q improvement can be used in a variety of ways but increased system reach beyond 1000 km and up to 3000 km has received the most interest since it has immediate commercial applicability and obviates the need for expensive electrical 3R regenerators. Unfortunately, the use of D-RAT to increase bandwidth is very difficult to implement effectively due to the Raman gain/tilt and other nonlinearities, all of which are discussed in U.S. Pat. No. 6,088,152 to Berger et al. To solve these problems, Berger et al. proposes preconditioning of the optical signals prior to transmitting the signals over an optical fiber span. Such preconditioning may either generate a gain that is pre-tilted opposite to the Raman gain/tilt, or may filter the Raman tilt/gain out of the received optical signals before the signals are re-transmitted.
While Berger et al. poses one possible solution to the Raman gain/tilt and nonlinearities, such preconditioning by generating an opposing gain or by filtering is also difficult and costly to implement. In addition, such preconditioning is not easily retrofittable and thus does not provide an easy and economical method for upgrading bandwidth in already installed fiber optic systems. Accordingly, there exists an unfulfilled need for a method for upgrading bandwidth in existing fiber optic systems which provides a substantial increase in the data carrying capacity without the disadvantages of the prior art systems and methods.